NPO publishes blog articles to inform and to stimulate conversation about issues of importance to NPO's mission.  All blog articles express the opinions of the authors as individuals and do not necessarily reflect the views of National Parents Organization, its Board of Directors, or its executives.  

February 5, 2018 by Robert Franklin, Esq, Member, National Board of Directors, National Parents Organization

The Warshak consensus report on overnights for very young children, based on 45 years of science on that issue and endorsed by 110 eminent scientists worldwide should have entirely put to rest the notion that those kids should spend all their time with their mothers. But the anti-dad crowd isn’t one to be daunted by empirical evidence scrupulously assembled and vetted. They have two studies, one by McIntosh, et al and one by Samantha Tornello with which to attempt to refute those 45 years of research and convince experts testifying in court that overnights for kids constitute a detriment to their well-being.

The Warshak Consensus Report identified significant problems and limitations in both studies that should affect the admissibility and weight of testimony that relies on these studies. As the U.S. Supreme Court in General Electric Co. v. Joiner noted: “[C]onclusions and methodology are not entirely distinct from one another. . . . A court may conclude that there is simply too great an analytical gap between the data and the opinion proffered.”

Stated another way, those two studies were so badly done that any expert opinion offered in court should be deemed inadmissible by the judge as failing to conform to the standard set for expert opinion evidence. The McIntosh and Tornello studies should never be allowed to sway judges or juries.

Why? For starters, the populations they studied bear little-to-no relation to the vast majority of litigants.

The Australian study’s sample of children under four years old is not representative of parents who are going through a divorce because most of the parents in the study were never married to each other (90% for the sample of infants and 71% for toddlers), and 41% had never even lived together.

So in many or even most cases, Dad may have been a stranger to the child. Therefore, spending overnights could well have caused confusion or anxiety in a very young child. But of course that’s seldom the case when a father seeks custody and parenting time.

The Tornello study’s population was even less like the typical divorcing couple.

The study’s data came from the Fragile Families sample of inner-city children born in impoverished circumstances: 62% of the age 1 sample lived below the poverty line, 60% of the parents were imprisoned before the children’s fifth birthdays, 85% were Black or Hispanic, 65% had parents who had non-marital births from more than one partner in their teenage or young adult years, and nearly two-thirds had not completed high school.

So much for the good news. Indeed, as different from the usual divorce litigant as those studies’ cohorts were, that’s the least of their problems.

For example, McIntosh, et al seem to have simply misrepresented their own findings, an odd approach to both social science and advocacy. The study states at the outset that,

“Infants under two years of age living with a nonresident parent for only one or more nights a week were more irritable and were more watchful and wary of separation from their primary caregiver than those primarily in the care of one parent.”


Only in the Appendix of the 169-page report can readers discover that the irritability score for babies with no overnights actually is slightly worse than the score for babies who spent one or more nights per week with their other parent. Also, the mean irritability score for the frequent overnighters and the infants in intact families was identical, and the mean irritability score for all groups was within the normal range.

The statement and the data in the appendix together come close to misrepresentation. The former refers to overall irritability and the latter to mean irritability, but at the very least the two have to be explicitly reconciled.

Then there’s the problem of sample size.

The irritability scores for infants with occasional overnights came from a sample of 14 infants. Only 11 infants saw their fathers on a schedule that would fit standard definitions of shared parenting. The sample sizes for the 2- to 3-year-olds with frequent overnights ranged from 5–25 depending on the variable analyzed (e.g., only five toddlers were rated for how well they got along with teachers and daycare attendants).

The best that can be said about such studies is that, if they had no other shortcomings, they might indicate the need for further research with larger populations. But the study was very badly done apart from its samples. As such, those samples render it essentially worthless.

But the flaws just keep coming.

The synopsis concluded that the overnighting infants were more “watchful and wary of separation from the primary caregiver.” The implication is that overnighting had somehow damaged the security of the babies’ relationships with their mothers. This conclusion, repeatedly cited to discourage overnights for children younger than two years of age, came from three questions that the researchers extracted from a standardized scale designed to measure young children’s readiness to learn language. The three questions are unreliable in the sense that they have not been established as a valid or reliable measure of children’s stress, anxiety, or attachments to their mother.

Yes, McIntosh, et al simply used a scale that’s been validated for a child’s readiness for language and decided that it was just as good a measure of something completely different and for which it’s never been validated. Think about that. Think about the level of frank dishonesty it would require to resort to such a tactic.

The apparently conscious misuse of the language precursor scale wasn’t the only such problem with the McIntosh study.

The Warshak Consensus Report observed that none of the four significant outcomes reported by McIntosh et al. were derived from measures that met basic scientific standards, a point also noted by Linda Nielsen in greater detail.

That’s 0 for 4. When a baseball player goes 0 for 4, he hopes to improve in the next game. It doesn’t tarnish his reputation, he just had a bad game of the type the best of the best sometimes have. Science is different. When a researcher does work that bad and publishes it for all the world to see, it doesn’t take long before other researchers start to mistrust her.

I’ll move on to the Tornello study next time.




National Parents Organization is a Shared Parenting Organization

National Parents Organization is a non-profit that educates the public, families, educators, and legislators about the importance of shared parenting and how it can reduce conflict in children, parents, and extended families. Along with Shared Parenting we advocate for fair Child Support and Alimony Legislation. Want to get involved?  Here’s how:

Together, we can drive home the family, child development, social and national benefits of shared parenting, and fair child support and alimony. Thank you for your activism.

#child'sbestinterests, #sharedparenting, #RichardWarshak

Share this post

Submit to FacebookSubmit to Google PlusSubmit to TwitterSubmit to LinkedIn